mm-intelligence/auburn-civic-data

Auburn Civic Data

This dataset packages public City of Auburn, Alabama civic GIS data for Ask Auburn and downstream civic-data analysis.

• Hugging Face repo: mm-intelligence/auburn-civic-data
• Generated: 2026-05-11 18:44 UTC
• Tables: 17
• Rows: 101,935
• Source: City of Auburn public ArcGIS services (data-coa.opendata.arcgis.com and gis.auburnalabama.org)

Tables

Table	Rows	Geometry	Description
city_limits	3	polygon	Auburn corporate-limits boundary polygons. Useful as a spatial filter ("is this point inside Auburn?") rather than as a query target on its own. Includes inside-limits and police-jurisdiction polygons.
wards	8	polygon	Auburn's 8 voting ward boundaries with the assigned council member and voting (polling) location for each ward. Used for point-in-polygon ward resolution ("which ward is this address in?") and ward-level aggregations. Auburn calls these "wards" — never "council districts" — though some legacy code paths still use the field name council_districts for the ward count in city context.
zip_codes	14	polygon	ZIP code boundary polygons covering the Auburn, Alabama area. ~14 polygons, used as a spatial reference layer for enriching other point-tables with ZIP via ST_Within. Includes ZIPs that extend outside the City of Auburn limits (portions of Opelika and Lee County).
parcels	33,050	polygon	Real-property parcel polygons for the City of Auburn (and immediate vicinity served by Auburn GIS). Each row represents one tax parcel with owner name, deed reference, subdivision, and acreage. ~33,050 polygons. The most useful table for property-level questions.
addresses	48,314	point	Street-address points covering the City of Auburn (and adjacent areas served by Auburn GIS). ~48,314 points. One row per addressable location with street name parts split out (pre-direction, name, post-type, etc.) and a full_address convenience field.
zoning	1,815	polygon	Zoning district boundaries for the City of Auburn. Defines land-use regulations (residential, commercial, industrial, mixed-use, etc.). ~1,815 polygons. Primarily a spatial reference layer — join other tables (parcels, addresses) TO this one via ST_Within rather than querying it alone.
zoning_overlay	107	polygon	Zoning OVERLAY district polygons — a second layer on top of the base zoning that adds extra rules (historic district, corridor, university neighborhood overlay, etc.). ~107 polygons. Always interpret in combination with the base auburn.zoning table.
streets	6,935	polyline	Centerline polylines for every public and private street in Auburn, including classification, surface type, lane counts, speed limit, and notes on whether each segment has a sidewalk or bike facility. ~6,935 polylines.
schools	14	point	Auburn City Schools campus locations (point markers). 14 schools. Pair with auburn.school_property for the campus footprint polygons.
school_property	38	polygon	Polygon footprints of school-owned campus property. 38 polygons. Useful for "is this address on a school campus?" lookups.
flood_plain	953	polygon	FEMA-derived flood-plain polygons covering Auburn. ~953 polygons. Includes flood zone designation (AE, X, etc.), base flood elevation, depth, and whether the area is inside a Special Flood Hazard Area.
fire_stations	13	point	Auburn Fire Division station locations. 13 stations covering the Auburn metro. Use for "where's the nearest fire station" queries and ward-level aggregation.
fire_zones	47	polygon	Auburn Fire Division response-zone polygons. ~47 zones used for dispatch and pre-planning. Pair with auburn.hydrants (each hydrant has a fire_zone field) and use ST_Within for "what fire zone covers my address?" lookups.
police_beats	14	polygon	Auburn Police Division patrol-beat boundaries. 14 beats covering the city. Use for "what beat covers my address?" and ward-level safety question routing.
hydrants	3,653	point	Auburn Water Works fire-hydrant inventory with location, ownership, install year, last static-pressure reading, last GPM flow, and associated fire zone. ~3,653 hydrants. Mostly useful for fire department / water utility queries; not a typical resident lookup.
garbage_routes	6,952	polyline	Per-street polylines tagged with the day of the week garbage gets picked up. ~6,952 segments covering Auburn's curbside service. For "when's my trash day?" lookups, find the nearest segment to the user's address (ST_DWithin within ~50m) — NOT a polygon point-in-polygon, since these are street centerlines.
polling_places	5	point	Auburn voting locations with address, hours, accessibility, contact info, the next election date, and the voter-registration deadline. 5 polling places. Use for "where do I vote?" queries.

Geometry

PostGIS geometry is exported as geometry_wkt. Point tables also include longitude and latitude derived from the source geometry. Polygon and line tables keep geometry as WKT only. No source nulls are manually filled.

Known Limitations

• Polling-place next election dates and voter-registration deadlines are not populated by the source layer.
• Some garbage route segments do not include a service_day value; these are preserved as nulls.
• Many street segments without street_name are driveways, private roads, Opelika-adjacent segments, or unclassified GIS features.
• The raw hydrants.fire_zone source field is almost empty; export includes derived fire_zone_spatial from Auburn fire-zone polygons.
• A few source records have null geometry and are preserved rather than inferred.

Public Data Notes

Parcel owner names are preserved because they are present in the official public GIS source. Users should verify critical decisions against the City of Auburn or Lee County source systems.

Loading

Each table is stored as a separate Parquet file:

from datasets import load_dataset

parcels = load_dataset("mm-intelligence/auburn-civic-data", "parcels")
addresses = load_dataset("mm-intelligence/auburn-civic-data", "addresses")

Direct Parquet loading also works:

import pandas as pd

parcels_df = pd.read_parquet("hf://datasets/mm-intelligence/auburn-civic-data/parcels/data.parquet")

Column-level quality metadata is available in data_quality.json and each table's column_quality.json.